The last model we consider is Credit Portfolio View (CPV), published by the consulting firm McKinsey in 1997. The focus of this top-down model is on the effect of macroeconomic factors on portfolio credit risk.
This approach models loss distributions from the number and size of credits in subportfolios, typically consisting of customer segments. Instead of considering fixed transition probabilities, this model conditions the default rate on the state of the economy, the assumption being that default rates increase during recessions. The default rate pt at time t is driven by a set of macroeconomic variables xk for various countries and industries through a linear combination called yt. It functional relationship to yt, called Jogit model, ensures that the probability is always between zero and one pt = 1/[1 + exp(yt)], yt = a + X Vkxk (23.13)
Using a multifactor model, each debtor is assigned to a country, industry, and rating segment. Uncertainty in recovery rates is also factored in. The model uses numerical simulations to construct the distribution of default losses for the portfolio. While useful for modeling default probabilities conditioned on the state of the economy, this approach is mainly top-down and does not generate sufficient detail of credit risk for corporate portfolios.
Was this article helpful?